“free float” assembly setup - sztaki · photogram-metry theodolite laser scanning architektural...
TRANSCRIPT
© WZL/Fraunhofer IPT
Reference Systems for a
“Free Float” Assembly Setup
IMEKO TC10 Workshop on Technical Diagnosis in the Cyber-Physical Era
Robert Schmitt WZL | RWTH Aachen Fraunhofer IPT
June 6th, 2017
Budapest, Hungary
Seite 2 © WZL/Fraunhofer IPT
I – Will Production remain the Same?
Seite 3 © WZL/Fraunhofer IPT
I – Will Production remain the Same?
II – Can Metrology contribute to
Tomorrow’s Production?
Seite 4 © WZL/Fraunhofer IPT
I – Will Production remain the Same?
II – Can Metrology contribute to
Tomorrow’s Production?
III – What should we assimilate in
Tomorrow’s Metrology?
Seite 5 © WZL/Fraunhofer IPT
Descriptive
Seite 6 © WZL/Fraunhofer IPT
Descriptive -> Diagnostic
Seite 7 © WZL/Fraunhofer IPT
Descriptive -> Diagnostic -> Predictive
Seite 8 © WZL/Fraunhofer IPT
Descriptive -> Diagnostic -> Predictive
Prescriptive
Seite 9 © WZL/Fraunhofer IPT
Experience
Observation:
Experience needs Expertise and Exercise
Training / Exercise
Talent /
Expertise
Seite 10 © WZL/Fraunhofer IPT
Observation:
The World becomes a Lab
Model
significance
(empirical and
phenomenological
evidence,
veracity / value)
availability of experiments (volume / variety)
Heuristic / probabilistic
limited significance
Seite 11 © WZL/Fraunhofer IPT
Local Lab of Labs
Observation:
The World becomes a Lab
Model
significance
(empirical and
phenomenological
evidence,
veracity / value)
availability of experiments (volume / variety)
Heuristic / probabilistic
limited significance
Seite 12 © WZL/Fraunhofer IPT
Local Lab of Labs
Observation:
The World becomes a Lab
Model
significance
(empirical and
phenomenological
evidence,
veracity / value)
availability of experiments (volume / variety)
Distributed Lab of Labs
Heuristic / probabilistic
limited significance
Empirical and phenomenological significance
Seite 13 © WZL/Fraunhofer IPT
I – A trial on Ontology
Seite 14 © WZL/Fraunhofer IPT
Measurement Systems Model Based Understanding
of Production
Human-Machine-Interaction New Production Paradigms
Seite 15 © WZL/Fraunhofer IPT
Measurement Systems Model Based Understanding
of Production
Human-Machine-Interaction New Production Paradigms
Seite 16 © WZL/Fraunhofer IPT
Receiving Data
Seite 17 © WZL/Fraunhofer IPT
industrial
metrology
Interferometry
industrial
photogram-
metry
theodolite
laser scanning
architektural and
technichal
photogrammetry
tachy-
metric
surveying
Franceschini, F.: Distributed Large-Scale Dimensional Metrology, Springer, London, 2011, S.4, ISO 10360
Differential
Global
Positioning
System
aerial photo-
grammetry
GPS
remote
sensing
Relationship between Object Size and „Accuracy“ of Measurement Methods
Object Size in mm
Ma
xim
um
Pe
rmis
sa
ble
Err
or
(MP
E)
in m
m
103
102
101
1
10−1
10−3
10−2
104
10−1 1 101 102 103 104 105
Seite 18 © WZL/Fraunhofer IPT
1- Innovative 1D Coordinate Measuring System for 3D Measurements
Intersecting Planes Technique (InPlanT)
Source:
Balsamo A, Egidi A, Francese C, Pisani M, 1D measurement of coordinates in space: a novel apparatus, euspen 2016, May 2016, Nottingham, UK
Pisani M, Balsamo A, Francese C, Cartesian approach to large scale co-ordinate measurement: InPlanT, LMPMI, September 2014, Tsukuba, Japan
Innovative measuring system working over 10m x 10m x 5m
Principle:
Prototype:
European LUMINAR Project
(Large volume Unified Metology for Industry)
Seite 19 © WZL/Fraunhofer IPT
2 - Absolute Interferometry
Divergent beam and Long range FSI
Source: Campbell M, Hughes B, A high-accuracy, self-calibrating and traceable coordinate measurement system, euspen 2016, May 2016, Nottingham, UK
Photodiode
Fourier
transform
Tuneable
frequency
laser, n
Reflectors Diverging
lens
Optical fibres
Innovative measuring system working over
10m x 10m x 5m
Seite 20 © WZL/Fraunhofer IPT
Measurement Systems Model Based Understanding
of Production
Human-Machine-Interaction New Production Paradigms
Seite 21 © WZL/Fraunhofer IPT
Source: FCMD Hattingen,
Model Based
Understanding
of Production
Seite 22 © WZL/Fraunhofer IPT
Refractive index compensation for LSM applications
Light-based Measurement Systems and Temperature
General assume that light travels in straight lines but thermal effects in particular cause light rays to
bend.
This could be a problem in factories, e.g. where air is cold on the floor and warm under the roof.
Source:
Robson S, Kyle S, MacDonald L, Shortis M, Towards understanding photogrammetric refraction in large volume metrology, LVMC, November 2014, Manchester, UK
Robson S, MacDonald L, Kyle S, Shortis M, Close range calibration of long focal length lenses in a changing environment, ISPRS 2016, July 2016, Prague, Czech Republic
Nuclear fuel assemblies Modelling of human movements for
computer games
Car body inspection
Seite 23 © WZL/Fraunhofer IPT
Refractive index compensation for LSM applications
Struggeling with Fermat’s Principle in Optical Systems
„low optical density“
„high optical density“
Seite 24 © WZL/Fraunhofer IPT
Refractive index compensation for LSM applications
Why we need the exact refractive index…
– laser wavelength = vacuum wavelength/refractive index
– “School” air ref. index = 1
– “University” air ref. index = 1.000 3
– “NMI” air ref. index = 1.000 276 (15 °C and 1 atm)
– = 1.000 266 (25 °C and 1 atm)
Source:
Robson S, MacDonald L, Kyle S, Shortis M, Close range calibration of long focal length lenses in a changing environment, ISPRS 2016, July 2016, Prague, Czech Republic
John A. Smith, CIRES, University of Colorado at Boulder
Fermat‘s Princple
Distance in m
Displacement in µm
Source: PTB
Seite 25 © WZL/Fraunhofer IPT
Model-based Refractive Index Compensation
Refraction calculation tool to simulate
images affected by refraction.
Reducing possible refraction errors in multi-
image tracking.
Source:
Robson S, Kyle S, MacDonald L, Shortis M, Boehm, J. Multi-camera systems for dimensional control in factories, EPMC, November 2015, Manchester, UK
Robson S, MacDonald L, Kyle S, Shortis M, Close range calibration of long focal length lenses in a changing environment, ISPRS 2016, July 2016, Prague, Czech Republic
Meiners-Hagen K, Pollinger F, Prellinger G, Rost K, Wendt K, Pöschel W, Dontsov D, Schott W, Mandryka V, Refractivity compensated tracking interferometer for precision engineering, IWK
Ilmenau, September 2014, Ilmenau, Germany
Refractive index compensation by two
colour interferometry
IFM mode: frequency doubled Nd: YAG laser (1064
nm+532 nm)
IFM measurement 1d uncertainty
U = (1µ𝑚)2+(10−7𝑙)2
Photogrammetry Interferometry
Seite 26 © WZL/Fraunhofer IPT
Models for workpieces under thermal and gravitational load (1/2)
EMRP Project ENG 56 “Drivetrain”
Traceable measurement of drive train
components for renewable energy
systems
Sources:
M. Stein, K. Kniel, F. Härtig: Reliable Measurements of Large Gears, presented at AGMA FTM 2014, October 2014 USA
M. Deni, G. Picotto: Traceable Measurements of Drivetrain Components for Renewable Energy Systems, presented at Gear Forum, March 2015, Italy
R. Schmitt, M. Peterek, Traceability in Large-Scale Metrology - Modeling Thermal Effects of Large DriveTrain Components, presented at EPMC European Portable Metrology
Conference, November 2015, UK
Involute gear standard and extra stiff coupling for measuring process Flatness measurement carried out on the upper reference plane of
a ring gear 3-point support (left), 4-point support (right)
Source: PTB
Seite 27 © WZL/Fraunhofer IPT
Computation
Table
WZL
RWTH Aachen
Computation
Table
Models for workpieces under thermal and gravitational load (2/2)
Workpiece, initial temperature
Ambient temperature curve
Material properties
Thermalisation time
Approximate temp. distribution (t)
Approximate, feature-specific
thermal expansion and displacement (t)
Computation
Table
Source:
R. Schmitt, M. Peterek, Traceability in Large-Scale Metrology - Modeling Thermal Effects of Large DriveTrain Components, presented at EPMC European Portable Metrology Conference,
November 2015, UK
General steps for model based analysis from temperature
data to displacement prediction for thermal compensation
Source: NPL
Source: WZL
Seite 28 © WZL/Fraunhofer IPT
CPPS
Model based Processes or CPPS
D3 Prüfstand
New Production Paradigms
Measurement Systems Model Based Understanding
of Production
Human-Machine-Interaction
Seite 29 © WZL/Fraunhofer IPT
CPPS
Model based Processes or CPPS
D3 Prüfstand
New Production
Paradigms
Seite 30 © WZL/Fraunhofer IPT
CPPS as a solution to the main challenges in Assembly
CPPS …
consist of autonomous and cooperative elements and subsystems that are getting into
connection with each other in situation dependent ways, on and across all levels of production,
from processes, through machines and production systems, up to production and logistics
networks…
• directly acquire physical data by using sensors and act on the physical world by using
actors
• analyse and store the acquired data and interact both with the physical and the virtual
world
• are networked amongst each other and within global information systems by wired or wireless
communication means
• use worldwide available data and services,
• dispose of several multi-modal human-machine-interfaces.
Models for measurement-systems and workpiece allow the prediction of the
interaction between both in the virtual measurement process (VMP)!
Seite 31 © WZL/Fraunhofer IPT
Virtual instruments and virtual measurement systems
Sources:
Muralikrishnan, B., Phillips, S., Sawyer, D., Laser Trackers for Large Scale Dimensional Metrology: A Review, Precision Engineering 44 13 (2016)
Ulrich T, Uncertainty Estimation for Kinematic Laser Tracker Measurements, IEEE Explore, 2015
Task
Automatic determination of the measurement uncertainty by simulation (numerical experiment)
with the virtual Coordinate Measuring Machine (vCMM).
Adaption of the principle for the virtual Laser Tracker (vLT)
Principle
• modelling the measuring
process
• determination of
- machine characteristic
- measuring conditions
- workpiece
• collecting and distorting
measuring points by Monte
Carlo simulation
• evaluating several up to
thousand measuring results
• calculating the
measurement uncertainty
Large-scale parts: Modelling of thermal
and gravitational behaviour is important!
Source: PTB
Source: PTB
LaserTracker and reference wall at PTB in
Braunschweig
Priciple of vCMM
Seite 32 © WZL/Fraunhofer IPT
Source: FFT, PAG
Lasertracker 6-DOF
Reference
Reflectors
Force-Torque
„Calibration“
Artefact
Sensor
Source: FFT, PAG
Multi Sensor Architecture
Towards a Metrology based Reference System
Seite 33 © WZL/Fraunhofer IPT
Assembly Robot
Basic Measurement System
Lasertracker (LT)
Airplane Shell
Force sensitive Jig
iGPS
Positioning Robot
Force Sensor (FT)
Use Case: Robot Assisted Adjustment of Airplane Shell
Communication network for system entities
Seite 34 © WZL/Fraunhofer IPT
GRS
iGPS-COS
Shell-COS
GRS H iGPS
GRS H Base KR60
GRS H Shell
KR60-Base
KR60-Tool
BaseH Tool
LT-Target
Use Case: Robot Assisted Adjustment of Airplane Shell
Different Local coordinate systems
Seite 35 © WZL/Fraunhofer IPT
GRS
GRS H Base KR60
KR60-Base
KR60-Tool
BaseH Tool
Bearing 1
Bearing 2
Bearing 3
Bearing 4
FB1
FB2
FB3
FB4
F_Robot
GRS H Bearing 3
GRS H Bearing 2
GRS H Bearing 1
Beam applied with
Strain Gauge
6D Force-/Torque
Sensor
Use Case: Robot Assisted Adjustment of Airplane Shell
Force Vectors
Seite 36 © WZL/Fraunhofer IPT
Use Case: Robot Assisted Adjustment of Airplane Shell
Set-up Messsystem
Lasertracker
FT Sensors Positioning
Robot 3
Positioning
Robot 2
Positioning
Robot 1
Seite 37 © WZL/Fraunhofer IPT
Cyberphysical Production Systems
Linking Real and Virtual World for Large-Scale Assembly
Messsystem
Lasertracker
Positionier-
roboter 3
Kraft-/ Momenten
Sensoren
Positionier-
roboter 2
Positionier-
roboter 1
„Real World“ „Virtual World“
Data
• Position
• Geometry
• Forces
Parameter
• Position for
robots
• Valid forces
• Path
planning
Sensors
Lasertracker
FT-Sensor
iGPS
Simulation and Control
Model-based analysis of data
Informationupdate
Modell
• MSA
• Parameters from real
world data
Fördergeber:
Seite 38 © WZL/Fraunhofer IPT
Measurement Systems Model Based Understanding
of Production
Human-Machine-Interaction New Production Paradigms
Seite 39 © WZL/Fraunhofer IPT
Where to find the
„smart expert“?
Seite 40 © WZL/Fraunhofer IPT
Example Process: Assembly of Truck Windscreen and Cabin
Navigation
e. g. UWB
Inspection,
e. g. Laser-
Tracker
Assembly
e. g. iGPS
The truck cabin is transported on a mobile platform
Low requirements regarding measurement uncertainty, but data needs to
be permanently available
The windscreen is mounted to the truck cabin
The measurement uncertainty must be suffiently low in order to avoid a
tolerance mismatch of the overall process
Optionally critical sections of the cabin can be checked after the assembly
process
The measurement uncertainty needs to be significantly lower (1 order of
magnitude) than the tolerance to allow reliable inspection
Seite 41 © WZL/Fraunhofer IPT
Free Float Flawless Assembly (F3A)
Assembly of Large Components in Motion
iGPS Transmitter
Controlled
Path
Planned Path
iGPS Receiver
Direction of Travel
A
Seite 42 © WZL/Fraunhofer IPT
Assembly of Large Components in Motion
Seite 43 © WZL/Fraunhofer IPT
II – Same Metrology as ever?
Seite 44 © WZL/Fraunhofer IPT
Self-stabilizing control loops no longer match production needs
Heuristic /
Determi-
nistic
Dynamic / instable Stationary / stable
(Pseudo-)
Chaotic /
Agile
Seite 45 © WZL/Fraunhofer IPT
Self-stabilizing control loops no longer match production needs
Heuristic /
Determi-
nistic
Dynamic / instable Stationary / stable
(Pseudo-)
Chaotic /
Agile
Seite 46 © WZL/Fraunhofer IPT
Self-stabilizing control loops no longer match production needs
(Pseudo-)
Chaotic /
Agile
Heuristic /
Determi-
nistic
Stationary / stable Dynamic / Instable
Seite 47 © WZL/Fraunhofer IPT
Self-stabilizing control loops no longer match production needs
Heuristic /
Determi-
nistic
Dynamic / Instable
Prescriptive
I II
III IV
(Pseudo-)
Chaotic /
Agile
Stationary / stable
Seite 48 © WZL/Fraunhofer IPT
Reference: Russel und Norvig 2010, D‘Andrea 2013
Research Area #1:
Learning from Data with a-priori-Knowledge
Seite 49 © WZL/Fraunhofer IPT
Reference: Levine (Google Research) 2016
Research Area #2:
Learning from meshed Systems
Seite 50 © WZL/Fraunhofer IPT
Research Area #3:
Interaction with AI-Agents
Seite 51 © WZL/Fraunhofer IPT
Research Area #4:
Empowering people in socio-technical Systems
Seite 52 © WZL/Fraunhofer IPT
III – Multi Agent Assembly:
Free Float, Flawless
Seite 53 © WZL/Fraunhofer IPT
‘‘Smart Manufacturing Systems”:
Sensor Based Metrological Frame
Seite 54 © WZL/Fraunhofer IPT
F3A: Free Float Flawless Assembly
An Industry 4.0 Vision for Future Manufacturing Systems
Real-Time Process
Feedback HMI
Product
Geometry
(high density
Point Cloud)
6DoF Actual Robot
Pose M2M
3D Navigation
Data M2M
6DoF Robot
Control M2M
Transfer and machine-readable
semantic description of machine-
data (control, measurements,
parameters) using OPC-UA
Wireless RT communication with
varying data rates (actual robot
pose via LaserTracker) to GB/s
for computer vision based quality
inspection
Model-based process control
– Descriptive: Global Reference
System to acquire current
system state
– Diagnostic: global process
model to track and control
– Predictive: Modelling mechanic
part behavior
– Prescriptive: Process control to
achieve quality goals
Seite 55 © WZL/Fraunhofer IPT
Global Reference Systems as a Ressource
© Automated Precision Inc., Pozyx Labs, Nikon AG
Cost & Availability
Measurement Uncertainty
Seite 56 © WZL/Fraunhofer IPT
Multi-Sensor Metrology Environment
Compatible Interfaces
Transformation of coordinates
and measurement uncertainty
Seite 57 © WZL/Fraunhofer IPT
Communication Architecture
OPC UA OPC UA
UDP
Sensor Sensor
Central Data
Processing Actors Mobile UI
Sensor
Seite 58 © WZL/Fraunhofer IPT
Integration Frameworks Connect All Relevant Entities to a Central
Control System and Allocate Resources Using OPC-UA
OPC-UA
IF
IF Device: iGPS
IF Device: AGV
IF Device: HMI
Device: Robot
Device: Robot.Model
OPC-UA
SOA-based
Control
System
IF Integration
Framework
Development Cycle Production Cycle User Cycle
Smart Expert
Middleware+
Correlation
AnalysisCluster
Algorithms
Smart Data
Multi-Modal Information Access
Data IntegrationData
ModelData Storage
and Caching
Analytics
Management of data access to proprietary systems
Digital Shadow
Aggregation and Synchronization
Learning
AlgorithmsMeta
Heuristics
Context sensitive
Processing
Application
Software
Raw Data Customer
Data
CRM
Machine
Data
MES
Product
Data
PLM
Test
Data
CAQ
Process
Data
ERP/
SCM
CAD
Data
CAD
Simulation
Data
FEM
Feedback
Data
BDE
Data Provision and Access
Decision Support
Agent Event-driven
Decisions A
Autonomous
Actions
Adaptive
Processes
Seite 59 © WZL/Fraunhofer IPT
Virtual Metrology Frame – Demo App @ WZL Lab
Seite 60 © WZL/Fraunhofer IPT
Side Track:
Global Reference not limited to Coordinates - Temperature
Wireless, magnetic Pt1000
Thermistors
Transmission via Bluetooth 4.1
Variable update rate
3D-printed housing
Transmission range of 15 m
Resolution of 20 mK
1 year battery lifetime @ 60Hz update rate
Water-proof, 0 °C to 80 °C operating range
Temperature data available via web-interface
35
mm
10
mm
Seite 61 © WZL/Fraunhofer IPT
Side Track:
3-D Live-Calculation of thermoelastic State
Seite 62 © WZL/Fraunhofer IPT
Unknown measurement uncertainties in production environments
Thermo-elastic monitoring of workpiece
Problem
Solution
Unknown temperature- &
deformation condition espc.
Large workpieces
Sensor- and model-based
temperature monitoring
for determination of deformation
Time dependant temperature distribution
and expected measurement uncertainty
for valid measurement results
Recommend minimum
tempering time
Ambient temperature Airflow Radiation
Coordinate meas. machine Laser radar
Laser tracker Measuring arm
Ambient and surface temperature
Seite 63 © WZL/Fraunhofer IPT
Big Picture: Data Processing for Agile Enterprises
Development Cycle Production Cycle User Cycle
Mobile
Apps and
Application
Software
Quality
Analytics Research Apps Predictive Maintenance
Raw Data
Machine
sensors
Test
data
Control
data
Simulation
data
ERP
Data
Work
planning
Product Lifecycle
Management CAQ
Data Data Envelope
Raw Data
Middleware+
Learning
Algorithms Analytics
Algorithms
Smart Data
Data Provision and Access (Batch and Streaming)
Aggregation and Synchronization
Multi-Modal Information Access
…and many more
Digital Shadowing Data Integration Data Storage
and Caching Model
Seite 64 © WZL/Fraunhofer IPT
0 Through Innovation in Consumer-Electronics,
Data is for free,
Information is precious.
Seite 65 © WZL/Fraunhofer IPT
I Against the background of shorter product life
cycles and increasing product complexity
production engineering will depend on
“more” and “better” understanding
Seite 66 © WZL/Fraunhofer IPT
II Getting more information for agile demands
from sophisticated methods for data analytics,
dynamic models
Seite 67 © WZL/Fraunhofer IPT
III “Smart expert” centered
socio-technical systems reduce unwanted
human induced errors, fault, flaws
by “Tangible Metrology”
Seite 68 © WZL/Fraunhofer IPT
Conclusion:
For tomorrow’s production, the ‘Digital Shadow’
sets a new paradigm in science, for both
profound understanding
and better estimation in an agile environment.
Casting the case-adjusted ‘Digital Shadow’ from
the ‘Digital Twin’ will be a metrological task
Seite 69 © WZL/Fraunhofer IPT
Thank You!
Seite 70 © WZL/Fraunhofer IPT
Thank You!
Seite 71 © WZL/Fraunhofer IPT
Gratias ago
Obrigado
Grazie
Merci
Thank you
Muchas
gracias
Vielen Dank
Eskerrik asko
Eυχαριστώ
많은 감사 Спасибо
Köszönöm Gràcies
Prof. Dr.-Ing. Robert H. Schmitt
Chair of Metrology and Quality Management
Laboratory for Machine Tools and Production Engineering WZL | RWTH Aachen University
Dankjewel